Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 11 2014, 10:25AMI was thinking about switching the stages in a multi-stage coil gun and I came up with the idea you can see in the schematic attached. It's a really messy schematic as I drew it very fast in paint. It's probably already been done if it works, but I haven't seen it and would like confirmation if it does work.
The triangle is the projectile and shows the direction of the projectile and the rectangles are the coils.
Update:
Or do I have to put R1 and R2 at the collector side of the PNP transistor?
Re: Multi-stage coil gun, switching the stages.
Yandersen, Sat Jan 11 2014, 12:00PM
Wow, have my respect for non-standard thinking! So the idea is "the coil must be ON when projectile dims the gate, but as soon as the next gate is dimmed, coil goes OFF no matter the condition of the gate". Good idea for coilguns with projectiles longer then the coils, but the schematic, even if would be drawn error-less, would not reliably work in practice unless the comparator with big hysteresis is used - otherwise with the same levels at + and - you will have noise at output. Unless resistors for positive feedback is added.
Besides, you do not need pnp transistor - simply switch positions of phototransistor and ballast resistor to invert the output.
Yandersen, Sat Jan 11 2014, 12:00PM
Wow, have my respect for non-standard thinking! So the idea is "the coil must be ON when projectile dims the gate, but as soon as the next gate is dimmed, coil goes OFF no matter the condition of the gate". Good idea for coilguns with projectiles longer then the coils, but the schematic, even if would be drawn error-less, would not reliably work in practice unless the comparator with big hysteresis is used - otherwise with the same levels at + and - you will have noise at output. Unless resistors for positive feedback is added.
Besides, you do not need pnp transistor - simply switch positions of phototransistor and ballast resistor to invert the output.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 11 2014, 01:20PM
Thx for your reply Yan, but shouldn't I keep the PNP transitors and feed them via a voltage divider so you would have different voltages when both fototransitors are off so the igbt's are still deactivated? or is this already fixed by the resistors for positive feedback. Also, would you just place those resistors over the output and the +? If I'm wrong, can you draw the schematic the way it might work? and maybe give a simple explanation? I'm sorry, but I don't know very much about electronics, especially not when it comes to comparators, but I can read something about them and see if I can learn some more about them.
Edit: Can't you use another voltage divider at the - of the comparator so it will get a slightly higher voltage than the + when they are both at 0? This way I don't think this positive feedback is neccesary, but maybe a really small positive feedback could still reduce the small amount of noise? Correct me if I'm wrong.
Dédé!, Sat Jan 11 2014, 01:20PM
Thx for your reply Yan, but shouldn't I keep the PNP transitors and feed them via a voltage divider so you would have different voltages when both fototransitors are off so the igbt's are still deactivated? or is this already fixed by the resistors for positive feedback. Also, would you just place those resistors over the output and the +? If I'm wrong, can you draw the schematic the way it might work? and maybe give a simple explanation? I'm sorry, but I don't know very much about electronics, especially not when it comes to comparators, but I can read something about them and see if I can learn some more about them.
Edit: Can't you use another voltage divider at the - of the comparator so it will get a slightly higher voltage than the + when they are both at 0? This way I don't think this positive feedback is neccesary, but maybe a really small positive feedback could still reduce the small amount of noise? Correct me if I'm wrong.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Sat Jan 11 2014, 02:48PM
Here is my recommendation (not schematic, cus I am lazy to draw it):
Phototransistor and ballast resistor (say, 1KOhm) are the same as in your schematic, and let's call the connection point between them as "control point".
Then, from the control point to + goes 100KOhm resistor. And from output to + goes 1MOhm resistor. The - goes directly to the control point of the next gate. Draw it nicely, post it.
The opamp works simple: if voltage level on + is higher then on -, output is high. Otherwise - low. Mention that there is no default value on inputs - left floating it may have any potential, resulting in unpredictable output behavior.
Yandersen, Sat Jan 11 2014, 02:48PM
Here is my recommendation (not schematic, cus I am lazy to draw it):
Phototransistor and ballast resistor (say, 1KOhm) are the same as in your schematic, and let's call the connection point between them as "control point".
Then, from the control point to + goes 100KOhm resistor. And from output to + goes 1MOhm resistor. The - goes directly to the control point of the next gate. Draw it nicely, post it.
The opamp works simple: if voltage level on + is higher then on -, output is high. Otherwise - low. Mention that there is no default value on inputs - left floating it may have any potential, resulting in unpredictable output behavior.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 11 2014, 03:47PM
What would be the voltage to supply to the circuit? 5V? I think 5V would be nice as it's commonly used and it's not far above the voltage the average LED takes.
Also I understand that part, but didn't fully understand how those resistors for feedback work. I think I do know after watching a video on youtube.
I'm currently drawing.
Dédé!, Sat Jan 11 2014, 03:47PM
What would be the voltage to supply to the circuit? 5V? I think 5V would be nice as it's commonly used and it's not far above the voltage the average LED takes.
Also I understand that part, but didn't fully understand how those resistors for feedback work. I think I do know after watching a video on youtube.
I'm currently drawing.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Sat Jan 11 2014, 03:53PM
5V is good for this part of the circuit, but too low to drive the IGBT, but you need a MOSFET driver IC for that purpose anyway - do you understand that driving the IGBT's gate by opamp is inappropriate due to low output current and slow speed of output change?
Anyway, there is no guarantee that the schematic will work as expected - need to be tested first.
Yandersen, Sat Jan 11 2014, 03:53PM
5V is good for this part of the circuit, but too low to drive the IGBT, but you need a MOSFET driver IC for that purpose anyway - do you understand that driving the IGBT's gate by opamp is inappropriate due to low output current and slow speed of output change?
Anyway, there is no guarantee that the schematic will work as expected - need to be tested first.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 11 2014, 04:13PM
Yes, I do, but what kind of mosfet driver ic do you mean and how would it be connected to the comparator's output?
Dédé!, Sat Jan 11 2014, 04:13PM
Yandersen wrote ...
do you understand that driving the IGBT's gate by opamp is inappropriate due to low output current and slow speed of output change?
do you understand that driving the IGBT's gate by opamp is inappropriate due to low output current and slow speed of output change?
Yes, I do, but what kind of mosfet driver ic do you mean and how would it be connected to the comparator's output?
Re: Multi-stage coil gun, switching the stages.
Yandersen, Sat Jan 11 2014, 04:53PM
Use non-inverting MOSFET driver and connect output of the opamp directly to the input of the driver. Simple as that. But first, you need to build 3-4 gates of the schematic we were discussing and check if it works at all - it may not. All I can say for sure is that yours was surely wrong. ^^
Yandersen, Sat Jan 11 2014, 04:53PM
Use non-inverting MOSFET driver and connect output of the opamp directly to the input of the driver. Simple as that. But first, you need to build 3-4 gates of the schematic we were discussing and check if it works at all - it may not. All I can say for sure is that yours was surely wrong. ^^
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 11 2014, 05:05PM
because of some reason I can't reply with an attachment.
Dédé!, Sat Jan 11 2014, 05:05PM
because of some reason I can't reply with an attachment.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 11 2014, 05:08PM
Here's the schematic. thx for your help btw even though it might not work (the first time)
Schematic
Btw, how do you do those green arrow links?
Edit: Would it be possible to make some kind of huge spear cannon this way which uses a very long and thin projectile? Of course you can still load it with normal, small ammo and you wouldn't have to make the entire spear ferromagnetic of course.
Dédé!, Sat Jan 11 2014, 05:08PM
Here's the schematic. thx for your help btw even though it might not work (the first time)
Schematic
Btw, how do you do those green arrow links?
Edit: Would it be possible to make some kind of huge spear cannon this way which uses a very long and thin projectile? Of course you can still load it with normal, small ammo and you wouldn't have to make the entire spear ferromagnetic of course.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Sat Jan 11 2014, 08:54PM
Here is how the link should be typed except that round braces must be rectangular:
(link=http://i39.tinypic.com/ vx2c5c.png)This is the text to click on(/link)
Long projectiles suck. The pull force is not directly proportional to the length of the bullet. Doubling the length will not double the force. If you go extreme, double the length will not increase force measurable at all. But the time coil must be energized (and dissipate heat) depends on the time bullet passes the coil. The faster the better. This means that short projectiles mean high efficiency (but higher currents if the energy is the same).
I use projectiles 1.25-1.5 times longer than their diameter.
Long projectiles mean low speed, high pulse (m*v), low efficiency, mild currents.
But long projectiles are not as bad as long coils, so you still can be on a good track, though.
Yandersen, Sat Jan 11 2014, 08:54PM
Here is how the link should be typed except that round braces must be rectangular:
(link=http://i39.tinypic.com/ vx2c5c.png)This is the text to click on(/link)
Long projectiles suck. The pull force is not directly proportional to the length of the bullet. Doubling the length will not double the force. If you go extreme, double the length will not increase force measurable at all. But the time coil must be energized (and dissipate heat) depends on the time bullet passes the coil. The faster the better. This means that short projectiles mean high efficiency (but higher currents if the energy is the same).
I use projectiles 1.25-1.5 times longer than their diameter.
Long projectiles mean low speed, high pulse (m*v), low efficiency, mild currents.
But long projectiles are not as bad as long coils, so you still can be on a good track, though.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 11 2014, 11:20PM
Actually, I didn't think spear like projectiles would be better, I just thought they would be cool. they might be more aerodynamic and they have a greater inertia.
Edit: Hey Yan, shouldn't there be a small resistor between the control point and the - of the previous comparator and maybe decrease the resistance of the 100K resistor. maybe 5K and 20K? Also maybe a resistor from the first control point to ground? This way you know for sure that when the projectile passes the second fototransistor, the second stage actually switches on and you know for sure that the - of the first comparator is dominant when they both see the projectile. Or am I wrong? schematic
Edit 2: shouldn't there also be a small resistor between the fototransistor and the control point? Let's say 50 Ohms? This way there's always a small voltage in the circuit, even if there's light everywhere. you would have 0.2v at the - of every comparator and almost 0.05 volt at the + of every comparator. schematic
Dédé!, Sat Jan 11 2014, 11:20PM
Actually, I didn't think spear like projectiles would be better, I just thought they would be cool. they might be more aerodynamic and they have a greater inertia.
Edit: Hey Yan, shouldn't there be a small resistor between the control point and the - of the previous comparator and maybe decrease the resistance of the 100K resistor. maybe 5K and 20K? Also maybe a resistor from the first control point to ground? This way you know for sure that when the projectile passes the second fototransistor, the second stage actually switches on and you know for sure that the - of the first comparator is dominant when they both see the projectile. Or am I wrong? schematic
Edit 2: shouldn't there also be a small resistor between the fototransistor and the control point? Let's say 50 Ohms? This way there's always a small voltage in the circuit, even if there's light everywhere. you would have 0.2v at the - of every comparator and almost 0.05 volt at the + of every comparator. schematic
Re: Multi-stage coil gun, switching the stages.
Yandersen, Sun Jan 12 2014, 03:12PM
The inputs of comparators have couple pF capacitance and dozen pA leakage current. It is absolutely insignificant - so just connect that input directly to the control point (resistor in series will simply have no effect). If you want to replace 100K with 20K - it is OK, but 1M must also be replaced with 200K then. This gives 10% voltage hysteresis to ensure difference between inputs. Just build that thing "wires in air", connect LED with 10K resistor to the outputs of comparators and move the finger between gates to see if the thing works. 3-4 gates is enough to be sure. If any output LED will glow dim, this means output oscillates. We expect 1 LED glow solidly having 2 gates crossed at once. For that test, you do not need to build the output stage with MOSFET drivers, neither barrel and coils required - just an optogates glued with thermal glue to some board and wires all around. :)
Yandersen, Sun Jan 12 2014, 03:12PM
The inputs of comparators have couple pF capacitance and dozen pA leakage current. It is absolutely insignificant - so just connect that input directly to the control point (resistor in series will simply have no effect). If you want to replace 100K with 20K - it is OK, but 1M must also be replaced with 200K then. This gives 10% voltage hysteresis to ensure difference between inputs. Just build that thing "wires in air", connect LED with 10K resistor to the outputs of comparators and move the finger between gates to see if the thing works. 3-4 gates is enough to be sure. If any output LED will glow dim, this means output oscillates. We expect 1 LED glow solidly having 2 gates crossed at once. For that test, you do not need to build the output stage with MOSFET drivers, neither barrel and coils required - just an optogates glued with thermal glue to some board and wires all around. :)
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sun Jan 12 2014, 03:23PM
Well, I'll just build it then, I think i will do it friday.
Dédé!, Sun Jan 12 2014, 03:23PM
Well, I'll just build it then, I think i will do it friday.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Mon Jan 13 2014, 02:23PM
Hey Yan, I just tried out your idea (the second schematic) and it didn't work. The LED of the second stage just glowed continuously. I first thought I did something wrong, but I triple checked and everything is according to the second schematic. I can't try out my idea, because the guy at the electronics shop gave me 51K resistors instead of 51 ohm resistors. I'll try out my idea tomorrow and see the results.
Dédé!, Mon Jan 13 2014, 02:23PM
Hey Yan, I just tried out your idea (the second schematic) and it didn't work. The LED of the second stage just glowed continuously. I first thought I did something wrong, but I triple checked and everything is according to the second schematic. I can't try out my idea, because the guy at the electronics shop gave me 51K resistors instead of 51 ohm resistors. I'll try out my idea tomorrow and see the results.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Mon Jan 13 2014, 03:10PM
Wait a sec. I understand why. Sorry. Try the following: insert 200 Ohm resistor between 1K and phototransistor (each stage). Then - of each comparator goes to the point between 1K and 200 Ohm resistors now. Will you try?
P.S.: 200 Ohm could be replaced by a resistor in range [150-510] Ohm.
P.P.S: did you measured voltages at phototransistor's collector when it lighted and dimmed?
Yandersen, Mon Jan 13 2014, 03:10PM
Wait a sec. I understand why. Sorry. Try the following: insert 200 Ohm resistor between 1K and phototransistor (each stage). Then - of each comparator goes to the point between 1K and 200 Ohm resistors now. Will you try?
P.S.: 200 Ohm could be replaced by a resistor in range [150-510] Ohm.
P.P.S: did you measured voltages at phototransistor's collector when it lighted and dimmed?
Re: Multi-stage coil gun, switching the stages.
Dédé!, Mon Jan 13 2014, 04:24PM
well, I don't have 200Ohm resistors, but I'll get some tomorrow together with some 51Ohm resistors. (I think 51 Ohms is enough isn't it?). Also, I still think there should be a resistor between the control point and the - of the comparator. (Wouldn't there else be no voltage going to the + of the comparator, because it's actually a voltage divider and if you have a 20K or 100k path and a 0Ohm path, wouldn't there be no voltage at the + of the comparator?) So I would think you should put a resistor that's about four times as small as the 100K or 20K resistor between the control point and the - of the comparator. So you would have 1V across the + of that comparator and 4V across the - of the previous comparator if the fototransistor doesn't see light and you would have between 0.05V and 0.2V across the + and between 0.2V and 0.7V across the - of the previous comparator when the fototransistor does see light. This was my idea for the fourth schematic I made.
You know what, I'll try again this saturday. I can make a better test setup this friday and then I'll try everything out saturday. I'll hook up my bro's oscilloscope to two outputs and I'll hook up my multimeter to another output. this way I can test it with 3 stages so I'll have a good idea of what works. For the fototransistors and leds I use BPW96B and OPE5685. I'll get 2 (multiplex?) boards and drill a few 5mm holes (every 5cm?). I'll let them face eachother and slide a board in between and watch if something happens. If I can find a way it works, I'll tell you. If I can't find a way it works, I'll tell you too of course. (This is the schematic I'm going to test next.)
Dédé!, Mon Jan 13 2014, 04:24PM
well, I don't have 200Ohm resistors, but I'll get some tomorrow together with some 51Ohm resistors. (I think 51 Ohms is enough isn't it?). Also, I still think there should be a resistor between the control point and the - of the comparator. (Wouldn't there else be no voltage going to the + of the comparator, because it's actually a voltage divider and if you have a 20K or 100k path and a 0Ohm path, wouldn't there be no voltage at the + of the comparator?) So I would think you should put a resistor that's about four times as small as the 100K or 20K resistor between the control point and the - of the comparator. So you would have 1V across the + of that comparator and 4V across the - of the previous comparator if the fototransistor doesn't see light and you would have between 0.05V and 0.2V across the + and between 0.2V and 0.7V across the - of the previous comparator when the fototransistor does see light. This was my idea for the fourth schematic I made.
You know what, I'll try again this saturday. I can make a better test setup this friday and then I'll try everything out saturday. I'll hook up my bro's oscilloscope to two outputs and I'll hook up my multimeter to another output. this way I can test it with 3 stages so I'll have a good idea of what works. For the fototransistors and leds I use BPW96B and OPE5685. I'll get 2 (multiplex?) boards and drill a few 5mm holes (every 5cm?). I'll let them face eachother and slide a board in between and watch if something happens. If I can find a way it works, I'll tell you. If I can't find a way it works, I'll tell you too of course. (This is the schematic I'm going to test next.)
Re: Multi-stage coil gun, switching the stages.
Yandersen, Tue Jan 14 2014, 12:28AM
OK, I've been masturbated enough to draw a final schematic for you. It is simple, fast, and will surely work (because it features my favourite long-used FET drivers). Order a dozen of MCP14E3 (or MCP14E9 if E3 are not available anymore) and here is the simple solution with just 1 resistor per stage:
Those drivers are referred as "Double Inverting MOSFET driver with Enable". Single chip has two equal drivers in it. Each driver has inverting input and "enable" input.
If voltage on Enable pin is low, the output of the driver is forced to be low no matter what input wants. If Enable is high, then output is determined by the input - if input goes low, output switches to high; if input goes high, output switches to low. So here is what we are doing: we connect Enable to the control point of next gate. This way output is allowed to be high ONLY if next gate is lighted. If it is lighted, then output value depends on the input: if control point of THIS gate (connected to In) has high potential (when ph is lighted), then output is low. If it falls close to ground (gate dimmed) output swings high turning on the IGBT. Once the gate got lighted again (or next gate got dimmed) the driver closes the IGBT.
The main difference between MCP14E3 and MCP14E9 is that the old E3 is 2 times more powerful (4A peak output current) - it gives faster switching time for your IGBT. The Enable function is not very common among FET drivers, but I found it extremely useful, though. Inverted In and Enable is equal to logical AND element with one inverted input.
Mention that those drivers are certified for capacitive loads - that means no resistor required between output of the driver and IGBT/MOSFET gate. It doesn't mean that they will not burn if you SC their output with Vcc or Gnd, though - it just assumes that heat dissipation limitations are not exceeded. It will drive your IGBT directly just fine, but if IGBT will burn melting the gate - driver will go in smoke too.
Good luck with your project!
Yandersen, Tue Jan 14 2014, 12:28AM
OK, I've been masturbated enough to draw a final schematic for you. It is simple, fast, and will surely work (because it features my favourite long-used FET drivers). Order a dozen of MCP14E3 (or MCP14E9 if E3 are not available anymore) and here is the simple solution with just 1 resistor per stage:
Those drivers are referred as "Double Inverting MOSFET driver with Enable". Single chip has two equal drivers in it. Each driver has inverting input and "enable" input.
If voltage on Enable pin is low, the output of the driver is forced to be low no matter what input wants. If Enable is high, then output is determined by the input - if input goes low, output switches to high; if input goes high, output switches to low. So here is what we are doing: we connect Enable to the control point of next gate. This way output is allowed to be high ONLY if next gate is lighted. If it is lighted, then output value depends on the input: if control point of THIS gate (connected to In) has high potential (when ph is lighted), then output is low. If it falls close to ground (gate dimmed) output swings high turning on the IGBT. Once the gate got lighted again (or next gate got dimmed) the driver closes the IGBT.
The main difference between MCP14E3 and MCP14E9 is that the old E3 is 2 times more powerful (4A peak output current) - it gives faster switching time for your IGBT. The Enable function is not very common among FET drivers, but I found it extremely useful, though. Inverted In and Enable is equal to logical AND element with one inverted input.
Mention that those drivers are certified for capacitive loads - that means no resistor required between output of the driver and IGBT/MOSFET gate. It doesn't mean that they will not burn if you SC their output with Vcc or Gnd, though - it just assumes that heat dissipation limitations are not exceeded. It will drive your IGBT directly just fine, but if IGBT will burn melting the gate - driver will go in smoke too.
Good luck with your project!
Re: Multi-stage coil gun, switching the stages.
ZakWolf, Tue Jan 14 2014, 02:37AM
motivated ?
ZakWolf, Tue Jan 14 2014, 02:37AM
Yandersen wrote ...
"OK, I've been masturbated enough to draw a final schematic for you."
"OK, I've been masturbated enough to draw a final schematic for you."
motivated ?
Re: Multi-stage coil gun, switching the stages.
Yandersen, Tue Jan 14 2014, 03:20AM
Duh...
Yandersen, Tue Jan 14 2014, 03:20AM
Duh...Re: Multi-stage coil gun, switching the stages.
Dédé!, Tue Jan 14 2014, 02:38PM
Well, I'm convinced your idea will work, but I'll try my idea first this friday or saturday and if it doesn't work, I'll try your idea. if my idea works, I'll design a PCB for my idea, if it doesn't work, I'll try your idea and if it works, I'll use your circuit.
Edit: Yan, I think you made a small mistake in your schematic. The last enable pin should also be connected to a phototransistor, because else the last coil won't switch of if the projectile is longer than the barrel.
Dédé!, Tue Jan 14 2014, 02:38PM
Well, I'm convinced your idea will work, but I'll try my idea first this friday or saturday and if it doesn't work, I'll try your idea. if my idea works, I'll design a PCB for my idea, if it doesn't work, I'll try your idea and if it works, I'll use your circuit.
Edit: Yan, I think you made a small mistake in your schematic. The last enable pin should also be connected to a phototransistor, because else the last coil won't switch of if the projectile is longer than the barrel.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Tue Jan 14 2014, 08:30PM
Yes, you right. What was the MOSFET drivers and IGBT you was going to use initially, anyway? Because if it all powered from LiPo battery under 100V, then MOSFET is a better choice and any MOSFET driver is an acceptable solution.
Yandersen, Tue Jan 14 2014, 08:30PM
Yes, you right. What was the MOSFET drivers and IGBT you was going to use initially, anyway? Because if it all powered from LiPo battery under 100V, then MOSFET is a better choice and any MOSFET driver is an acceptable solution.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Tue Jan 14 2014, 09:42PM
no, it´s powered by 400V 16.8mF cap bank. IXYK140N90C3 was my choice for the IGBTs and IXDN614PI was my choice for the gate driver and I'm using uzzors' zvs cap charger, just so you know. I think I'll first make 4 stages and later on, I might make some more and also increase cap bank size. I'm planning to use your coil winding technique btw.
Dédé!, Tue Jan 14 2014, 09:42PM
no, it´s powered by 400V 16.8mF cap bank. IXYK140N90C3 was my choice for the IGBTs and IXDN614PI was my choice for the gate driver and I'm using uzzors' zvs cap charger, just so you know. I think I'll first make 4 stages and later on, I might make some more and also increase cap bank size. I'm planning to use your coil winding technique btw.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Tue Jan 14 2014, 10:47PM
Well, with MCP14E9 your switching time will be around 100ns, with IXDN614 your IGBT will swing in 10ns. Yes, the last one is better in general, but your gun does not switches millions times per second, so you will not feel any difference with any of those drivers. But the schematic is so much simpler with just one IC, isn't it? Consider switching takes 100ns, happens at current 800A and 800V voltage at collector (worst IMpossible case). The amount of energy dissipated will be:
E=U*I*t=64mJ
-per shot per IGBT. Is it really so enormous to care about?.. The amount of heat dissipated on collector-emitter drop during the ON period will be much higher than that anyway.
Yandersen, Tue Jan 14 2014, 10:47PM
Well, with MCP14E9 your switching time will be around 100ns, with IXDN614 your IGBT will swing in 10ns. Yes, the last one is better in general, but your gun does not switches millions times per second, so you will not feel any difference with any of those drivers. But the schematic is so much simpler with just one IC, isn't it? Consider switching takes 100ns, happens at current 800A and 800V voltage at collector (worst IMpossible case). The amount of energy dissipated will be:
E=U*I*t=64mJ
-per shot per IGBT. Is it really so enormous to care about?.. The amount of heat dissipated on collector-emitter drop during the ON period will be much higher than that anyway.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Wed Jan 15 2014, 12:12PM
I don't really care about simplicity. I would like it better if my gun has a circuit inside that's partially designed by me. If I can get my idea working, I'll use my idea; If I can't get my Idea working, but yours will work, I'll use your idea. I know those small differences in numbers don't matter and using a single IC and a lot less components is easier, but I just think it's nice to be able to say I made a contribution to the designing of the circuit I'm using in my coil gun. Also, it's not that much harder to build up my circuit, but I'll first have to see if it works or not.
Dédé!, Wed Jan 15 2014, 12:12PM
I don't really care about simplicity. I would like it better if my gun has a circuit inside that's partially designed by me. If I can get my idea working, I'll use my idea; If I can't get my Idea working, but yours will work, I'll use your idea. I know those small differences in numbers don't matter and using a single IC and a lot less components is easier, but I just think it's nice to be able to say I made a contribution to the designing of the circuit I'm using in my coil gun. Also, it's not that much harder to build up my circuit, but I'll first have to see if it works or not.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Wed Jan 15 2014, 09:55PM
Totally agree. I understand and deeply respect that. Howether, to make anything by yourself you need to learn a lot so to put parts not in a random fashion or according to intuition.
F.e. in your schematic the collectors go to the opamp's inputs, but there are no ballast resistors to pull the collectors down. You also need to think over all cases (combinations of lighted/dimmed gates) and model the voltage potentials in all possible cases. If any result in equal voltages at + and - of the opamp - the output is undefined and may even start oscillating quickly destructing the IGBT.
Yandersen, Wed Jan 15 2014, 09:55PM
Totally agree. I understand and deeply respect that. Howether, to make anything by yourself you need to learn a lot so to put parts not in a random fashion or according to intuition.
F.e. in your schematic the collectors go to the opamp's inputs, but there are no ballast resistors to pull the collectors down. You also need to think over all cases (combinations of lighted/dimmed gates) and model the voltage potentials in all possible cases. If any result in equal voltages at + and - of the opamp - the output is undefined and may even start oscillating quickly destructing the IGBT.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Thu Jan 16 2014, 05:45PM
you mean like a big resistor in parallel with the 50Ohm resistor and the fototrasistor? If so, I don't see why that would be necessary. Also, Isn't the resistor over the output and the + of the comparator supposed to prevent oscillation of the output and don't the resistance differences of the resistors connected to the inputs of the comparators and the control point and the 50Ohm connected to the collector prevent the same voltage at both input pins?
Dédé!, Thu Jan 16 2014, 05:45PM
you mean like a big resistor in parallel with the 50Ohm resistor and the fototrasistor? If so, I don't see why that would be necessary. Also, Isn't the resistor over the output and the + of the comparator supposed to prevent oscillation of the output and don't the resistance differences of the resistors connected to the inputs of the comparators and the control point and the 50Ohm connected to the collector prevent the same voltage at both input pins?
Re: Multi-stage coil gun, switching the stages.
Yandersen, Thu Jan 16 2014, 08:43PM
No, I mean the collector of the transistor - it can only pull up, and without the ballast resistor to pull that point down it will just stay there.
For some reason you think that + and - of the opamp pose any actions - they are not. Inputs of the opamp is like a probes of a multimeter - they measure whatever you give them. They neither pull measuring point up nor down. If you have collector of the pnp transistor pulling the input up, then be sure to add resistor to automatically pull points down when transistor goes off.
Yandersen, Thu Jan 16 2014, 08:43PM
No, I mean the collector of the transistor - it can only pull up, and without the ballast resistor to pull that point down it will just stay there.
For some reason you think that + and - of the opamp pose any actions - they are not. Inputs of the opamp is like a probes of a multimeter - they measure whatever you give them. They neither pull measuring point up nor down. If you have collector of the pnp transistor pulling the input up, then be sure to add resistor to automatically pull points down when transistor goes off.
Re: Multi-stage coil gun, switching the stages.
Dédé!, Sat Jan 25 2014, 01:04AM
Hey Yan, I've got the schematic (kind of) working. At least, with direct connections to the ground instead of fototransistors. I'm using my final schematic, but with a 180Ohm resistor instead of a 50Ohm connected to the collector (or in this case, a wire going to the ground). Tomorrow I'll test my schematic with the fototransistors. For now, I got around 4.5V at the output when the first wire isn't connected to ground anymore and the second is. all other configurations give less than half a volt at the output. I've tried it with 3 stages for now, tomorrow I'll test 3 stages again, but with light barriers. I haven't tried your idea, with MCP14E3's, but I will definitely try it, even if I can get my circuit working with fototransistors etc.
Dédé!, Sat Jan 25 2014, 01:04AM
Hey Yan, I've got the schematic (kind of) working. At least, with direct connections to the ground instead of fototransistors. I'm using my final schematic, but with a 180Ohm resistor instead of a 50Ohm connected to the collector (or in this case, a wire going to the ground). Tomorrow I'll test my schematic with the fototransistors. For now, I got around 4.5V at the output when the first wire isn't connected to ground anymore and the second is. all other configurations give less than half a volt at the output. I've tried it with 3 stages for now, tomorrow I'll test 3 stages again, but with light barriers. I haven't tried your idea, with MCP14E3's, but I will definitely try it, even if I can get my circuit working with fototransistors etc.
Re: Multi-stage coil gun, switching the stages.
Yandersen, Sat Jan 25 2014, 03:50AM
Oh, surprising. Well, whatever works better for you. Just test the circuit thoroughly - if any driver jams in ON state, IGBT will go in smokes, you know.
Yandersen, Sat Jan 25 2014, 03:50AM
Oh, surprising. Well, whatever works better for you. Just test the circuit thoroughly - if any driver jams in ON state, IGBT will go in smokes, you know.
Print this page